Spatial-temporal evolution prediction of train-induced settlement in railway transition zone using a simplified iterative framework

Abstract

Accurate prediction of train-induced settlement in railway transition zones is of paramount importance for ensuring the safety and serviceability of high-speed railway (HSR) infrastructure. The inherent complexity of mechanical properties and settlement distribution in these zones stems from the significant stiffness variation between different track structures. This study presents a novel iterative framework for long-term settlement prediction specifically tailored to ballastless track transition zones of HSR systems. The framework couples a dynamic Train-Track-Transition Zone (TTTZ) model with a plastic strain prediction model for soil, enhanced by a jump-step iterative algorithm that improves computational efficiency while maintaining accuracy. The model's validity has been verified through comprehensive comparisons with in-situ measurements and existing analytical solutions. Numerical results demonstrate that the iterative updating of track irregularities is crucial for accurate settlement prediction, as it accounts for the time-dependent dynamic characteristics of the TTTZ system. Furthermore, a wavelet transform-short energy method is developed to identify high-density vibration energy distributions in the spatial domain, establishing a robust correlation between dynamic responses and settlement evolution. This study underscores the importance of iterative modeling and advanced time-frequency analysis in settlement prediction and track quality assessment, offering valuable insights for the design, maintenance, and evaluation of HSR transition zones.